Order purchase and delivery method, system and device based on data analysis

Abstract

The invention discloses an order purchase and delivery method, system and device based on data analysis, and the method comprises the steps: obtaining supplier evaluation data of a target supplier, and carrying out the analysis, and obtaining supplier performance data; performing risk assessment on the supplier performance data based on a supplier analysis model to obtain supplier risk data; obtaining logistics evaluation data, and establishing a logistics data analysis model to analyze the logistics evaluation data to obtain logistics performance data; obtaining related information of a customer order, and obtaining an optimal supplier and an optimal logistics company corresponding to the customer order in combination with the supplier performance data, the supplier risk data and the logistics performance data; and based on the related information of the customer order, performing path optimization through a path selection algorithm to obtain an order delivery path, and completing delivery based on the order delivery path and the related information of the customer order. According to the invention, data analysis is carried out on suppliers and logistics companies, so that the order purchase and distribution efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data analysis, and in particular to an order purchasing and delivery method, system and device based on data analysis. Background Art

[0002] At present, when merchants make purchases, the purchasing department obtains the quotation provided by the current supplier and completes the purchase after approval. When the purchasing department receives a lower quotation than the current supplier, it needs to investigate the supplier with the lower quotation before making the purchase. In the case of asymmetric market information, when the midstream glass manufacturers and purchasing merchants do not have enough information, they cannot timely and accurately understand the price trend of the raw material market, changes in customer demand and other information, which makes it easy for merchants to make decisions that are not in their favor during the procurement process. Moreover, under the existing distribution model, multi-level distributors add prices layer by layer, and the purchase price of downstream merchants is relatively high. Since manufacturers cannot accurately obtain merchant orders, they can only formulate production plans based on experience, resulting in long inventory time, high costs, low turnover rates, and strong regionality. It cannot meet the increasingly diverse needs of merchants and requires more flexible and efficient solutions.

[0003] In the process of delivering purchased orders, the existing delivery methods do not conduct comprehensive analysis and screening of logistics companies, are unable to rate logistics companies, and cannot monitor the delivery quality of logistics companies in real time, resulting in low efficiency in the existing delivery methods and problems such as empty return trips, and are unable to guarantee timely and economical transportation of orders. Summary of the invention

[0004] In view of the shortcomings of the prior art, the present invention provides an order purchasing and delivery method, system and device based on data analysis.

[0005] In order to solve the above technical problems, the present invention is solved by the following technical solutions: An order purchasing and delivery method based on data analysis comprises the following steps: Obtain supplier evaluation data of target suppliers, build a supplier scoring model to analyze the supplier evaluation data, and obtain supplier performance data, where the supplier evaluation data includes quality data, price data, and delivery data; Establish a supplier analysis model, conduct risk assessment on supplier performance data based on the supplier analysis model, and obtain supplier risk data; Obtain logistics evaluation data of logistics companies, build a logistics data analysis model to analyze the logistics evaluation data, and obtain logistics performance data, where the logistics evaluation data includes timeliness data, safety data, cost data, service quality data and coverage data; Obtain relevant information about customer orders, and combine supplier performance data, supplier risk data, and logistics performance data to obtain the best supplier and best logistics company that match the customer order, wherein the relevant information about customer orders includes customer order data and customer address data; Based on customer address data, the initial order delivery path is obtained through the path selection algorithm and the initial order delivery path is optimized based on the path optimization algorithm to obtain the order delivery path, and the customer order delivery is completed based on the order delivery path and customer order data.

[0006] As an implementable method, the supplier rating model is constructed to analyze the supplier evaluation data to obtain supplier performance data, including the following steps: Obtain and analyze the target supplier's supply qualification rate data and customer evaluation data to obtain the target supplier's quality data, where the customer evaluation data includes after-sales service, customer evaluation and communication efficiency. The quality data is expressed as follows:

[0007] Obtain supplier quotation data and market average price and analyze them to obtain target supplier price data, which is expressed as follows:

[0008] Obtain and analyze the on-time delivery rate and emergency response capability data to obtain the target supplier's delivery data, which is expressed as follows:

[0009] A supplier scoring model is established based on quality data, price data and delivery data. The target supplier evaluation data is analyzed based on the supplier scoring model to obtain supplier performance data, which is expressed as follows:

[0010] in, Indicates quality data, Indicates the pass rate data, Represents customer evaluation data, Represents price data, Represents supplier quotation data, represents the average market price, Indicates delivery data, Indicates the on-time delivery rate, Represents emergency response capability data, represents supplier performance data, , , , Indicates the evaluation data weight, , , Indicates the weight of supplier performance data.

[0011] As an implementable method, the supplier analysis model includes one of a SVM classification model, a K-means clustering model and a Bayesian classification model; The SVM classification model is trained by fitting the supplier performance data and the corresponding real risk labels, and performs classification prediction based on the training results and the supplier performance data to obtain the supplier risk data; The K-means clustering model clusters the supplier performance data and analyzes the clustering results to obtain supplier risk data; The Bayesian classification model analyzes and predicts the real risk labels and prior probabilities of supplier performance data through Bayesian theorem to obtain supplier risk data.

[0012] As an implementable method, the SVM classification model is trained by fitting the supplier performance data and the corresponding real risk labels, and classification prediction is performed based on the training results and the supplier performance data to obtain the supplier risk data, including the following steps: Conduct risk assessment on supplier performance data and obtain the real risk label corresponding to the supplier performance data; Establish an SVM classification pre-training model and a risk assessment loss function, train the SVM classification pre-training model based on supplier performance data and the corresponding real risk labels, and optimize it based on the risk assessment loss function to obtain the SVM classification model; Based on the SVM classification model, the performance data of the supplier to be tested is risk assessed to obtain the supplier risk data; The risk assessment loss function is expressed as follows:

[0013] in, represents the risk assessment loss function, represents supplier performance data, Indicates the actual risk label corresponding to the supplier performance data, represents the regularization parameter, represents the normal vector of the SVM classification model, represents the bias of the SVM classification model, Indicates Supplier performance data, Represents the quantity of supplier performance data.

[0014] As an implementable method, the K-means clustering model clusters the supplier performance data, analyzes the clustering results, obtains outlier data and analyzes them, and obtains supplier risk data, including the following steps: The number of classification clusters and the classification cluster centers are preset, cluster allocation is performed based on the distance between the supplier performance data and the classification cluster centers, and the classification cluster centers after cluster allocation are calculated, wherein the distance between the supplier performance data and the classification cluster centers is calculated by the following formula:

[0015] The classification cluster center is calculated by the following formula:

[0016] Iterate the cluster allocation and the calculation of the classification cluster center until the classification cluster center no longer changes, obtain the performance data classification result, set the distance threshold, obtain the supplier performance data of the performance data classification result whose distance from the classification cluster center does not meet the distance threshold and analyze it to obtain the supplier risk data; in, represents the distance between the supplier performance data and the classification cluster center, represents supplier performance data, Indicates Supplier performance data, Indicates The cluster centers, represents the quantity of supplier performance data, Indicates the number of classification clusters.

[0017] As an implementable method, the Bayesian classification model analyzes and predicts the real risk labels and prior probabilities of supplier performance data by Bayesian theorem to obtain supplier risk data, including the following steps: Obtain the prior probability of supplier risk, analyze historical supplier performance data, and obtain the prior probability of historical supplier performance data and the prior probability of supplier performance when supplier risks occur; Based on the prior probability of supplier risk, the prior probability of historical supplier performance data corresponding to when the supplier has risk, and the prior probability of supplier performance data, the supplier performance data is analyzed to obtain the supplier risk category probability, which is expressed as follows:

[0018] The maximum value of the supplier risk category probability is selected as the supplier risk data, which is expressed as follows:

[0019] in, represents the supplier risk category probability, represents the prior probability of supplier risk, Represents the prior probability of historical supplier performance data corresponding to when a supplier is at risk, represents the prior probability of supplier performance data, Indicates Supplier performance data, represents supplier performance data, Indicates the number of supplier performance data. Represents supplier risk data, Represents historical supplier risk data.

[0020] As an implementable method, building a logistics data analysis model to analyze logistics evaluation data to obtain logistics performance data includes the following steps: Obtain the on-time delivery order data and total order data of the logistics company, and analyze them to obtain the timeliness data, which is shown as follows:

[0021] Get the damaged order data of the logistics company, analyze it in combination with the total order data, and get the safety data, which is shown as follows:

[0022] Obtain and analyze the quotation data of logistics companies and the average price data of logistics to obtain the cost data, which is shown as follows:

[0023] Analyze customer feedback data, service response data and tracking information data to obtain service quality data, and obtain coverage data through logistics network data and transportation coverage data; The timeliness data, safety data, cost data, service quality data and coverage data are standardized, and the standardized timeliness data, cost data, service quality data and coverage data are taken as the ideal solution, and the standardized safety data is taken as the negative ideal solution. The closeness of the logistics evaluation data is obtained, and the logistics performance data is obtained, which is expressed as follows:

[0024] in, Indicates time-sensitive data. Indicates on-time delivery of order data, Represents the total order data, Indicates safety data. Indicates broken order data, Represents cost data, Indicates the quotation data of logistics companies. Represents the average logistics price data, Represents logistics performance data, represents a negative ideal solution, represents an ideal solution.

[0025] As an implementable method, the optimization of the initial order delivery path includes the following steps: Analyze customer order data and customer address data, group customer orders with similar destinations for delivery, and perform return delivery based on customer order data; Establish a demand forecasting model, combine customer order data to predict customer order demand, obtain the priority ranking of customer order data, and dynamically adjust the initial order delivery path based on customer order demand and priority ranking; The initial order delivery path is optimized through the shortest path algorithm and vehicle path optimization algorithm to obtain the order delivery path.

[0026] As an implementable method, the optimization of the initial order delivery path by the shortest path algorithm includes the following steps: Get the transport node based on the customer address data, create a transport open list and a transport closed list based on the transport node, get the start node and the destination node through the transport node and add them to the transport open list; Establish a transportation heuristic function, select the transportation node that minimizes the transportation heuristic function from the open list as the current node, and add the current node to the closed list. Iterate until the current node is the target node to obtain the order delivery path, where the transportation heuristic function is expressed as follows:

[0027] in, represents the transport heuristic function, Indicates the actual cost from the starting node to the current node, represents the heuristic estimate of the cost from the current node to the target node, Indicates a transport node.

[0028] As an implementable method, the initial order delivery path is optimized by a vehicle path optimization algorithm, including the following steps: Based on the customer order data, the transportation node is obtained, the order transportation cost is obtained through the transportation node, and the node decision is obtained to analyze whether to pass through the current transportation node. The transportation objective function is established based on the order transportation cost and node decision, which is expressed as follows:

[0029] By adjusting the node decision and combining the order transportation cost between transportation nodes, the transportation objective function is minimized and the order delivery path is obtained; in, represents the transportation objective function, Indicates the transport node To transport node The shipping cost of orders between Indicates passing through a transport node To transport node Node decision making, Indicates the number of transport nodes.

[0030] As an implementable method, after completing the delivery of customer orders, the logistics performance data is updated through order delivery data, wherein the order delivery data includes empty load rate data, delivery time efficiency data and comprehensive route optimization data; The empty load rate data is obtained by obtaining the empty load rate of the best logistics companies in the process of customer order delivery and analyzing it; The delivery time efficiency data is obtained by real-time tracking of the time efficiency of the best logistics company during the delivery process and analyzing the time efficiency of reaching the transportation node during the transportation process; The route optimization data is obtained through a comprehensive analysis of the cost, distance and vehicle empty rate of each transportation route of the best logistics company.

[0031] An order procurement and distribution system based on data analysis, including a supplier evaluation module, a risk analysis module, a logistics evaluation module, an order confirmation module and an order distribution module; The supplier evaluation module obtains supplier evaluation data of the target supplier, builds a supplier scoring model to analyze the supplier evaluation data, and obtains supplier performance data, wherein the supplier evaluation data includes quality data, price data, and delivery data; The risk analysis module establishes a supplier analysis model, performs risk assessment on supplier performance data based on the supplier analysis model, and obtains supplier risk data; The logistics evaluation module obtains logistics evaluation data of the logistics company, builds a logistics data analysis model to analyze the logistics evaluation data, and obtains logistics performance data, wherein the logistics evaluation data includes timeliness data, safety data, cost data, service quality data and coverage data; The order determination module obtains relevant information of the customer order, combines supplier performance data, supplier risk data and logistics performance data, and obtains the best supplier and the best logistics company that match the customer order, wherein the relevant information of the customer order includes customer order data and customer address data; The order delivery module obtains the initial order delivery path based on the customer address data through a path selection algorithm and optimizes the initial order delivery path based on a path optimization algorithm to obtain the order delivery path, and completes the customer order delivery based on the order delivery path and the customer order data.

[0032] A computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the following method is implemented: Obtain supplier evaluation data of target suppliers, build a supplier scoring model to analyze the supplier evaluation data, and obtain supplier performance data, where the supplier evaluation data includes quality data, price data, and delivery data; Establish a supplier analysis model, conduct risk assessment on supplier performance data based on the supplier analysis model, and obtain supplier risk data; Obtain logistics evaluation data of logistics companies, build a logistics data analysis model to analyze the logistics evaluation data, and obtain logistics performance data, where the logistics evaluation data includes timeliness data, safety data, cost data, service quality data and coverage data; Obtain relevant information about customer orders, and combine supplier performance data, supplier risk data, and logistics performance data to obtain the best supplier and best logistics company that match the customer order, wherein the relevant information about customer orders includes customer order data and customer address data; Based on customer address data, the initial order delivery path is obtained through the path selection algorithm and the initial order delivery path is optimized based on the path optimization algorithm to obtain the order delivery path, and the customer order delivery is completed based on the order delivery path and customer order data.

[0033] An order purchasing and distribution device based on data analysis includes a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor implements the following method when executing the computer program: Obtain supplier evaluation data of target suppliers, build a supplier scoring model to analyze the supplier evaluation data, and obtain supplier performance data, where the supplier evaluation data includes quality data, price data, and delivery data; Establish a supplier analysis model, conduct risk assessment on supplier performance data based on the supplier analysis model, and obtain supplier risk data; Obtain logistics evaluation data of logistics companies, build a logistics data analysis model to analyze the logistics evaluation data, and obtain logistics performance data, where the logistics evaluation data includes timeliness data, safety data, cost data, service quality data and coverage data; Obtain relevant information about customer orders, and combine supplier performance data, supplier risk data, and logistics performance data to obtain the best supplier and best logistics company that match the customer order, wherein the relevant information about customer orders includes customer order data and customer address data; Based on customer address data, the initial order delivery path is obtained through the path selection algorithm and the initial order delivery path is optimized based on the path optimization algorithm to obtain the order delivery path, and the customer order delivery is completed based on the order delivery path and customer order data.

[0034] The present invention has significant technical effects due to the adoption of the above technical solution: The present invention obtains supplier performance data and logistics performance data by analyzing supplier evaluation data and logistics evaluation data respectively, obtains supplier risk data by performing risk analysis based on supplier performance data, obtains the best supplier and the best logistics company by analyzing supplier performance data, obtains order delivery data through relevant information of customer orders and performing path optimization, and completes the delivery of customer orders in combination with the best supplier and the best logistics company. By analyzing multiple information of suppliers and logistics companies, high-quality suppliers and logistics companies are selected for customers, and the management of the supply chain is optimized to achieve the purpose of improving corporate competitiveness and efficiency, respond to the increasingly fierce industry competition, diversified customer needs, increase supply chain complexity and reduce the pressure of cost control, meet the risk management needs of suppliers and achieve sustainable development. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0036] Figure 1 It is a schematic flow diagram of the method of the present invention; Figure 2 It is an overall schematic diagram of the system of the present invention; Figure 3 It is a schematic diagram of the process of the glass trading platform of the present invention. DETAILED DESCRIPTION

[0037] The present invention is further described in detail below in conjunction with embodiments. The following embodiments are for explanation of the present invention but the present invention is not limited to the following embodiments.

[0038] Embodiment 1: An order purchasing and delivery method based on data analysis, such as Figure 1 As shown, the following steps are included: S100, obtaining supplier evaluation data of target suppliers, building a supplier scoring model to analyze the supplier evaluation data, and obtaining supplier performance data, wherein the supplier evaluation data includes quality data, price data, and delivery data; S200, establishing a supplier analysis model, and performing risk assessment on supplier performance data based on the supplier analysis model to obtain supplier risk data; S300, obtaining logistics evaluation data of the logistics company, building a logistics data analysis model to analyze the logistics evaluation data, and obtaining logistics performance data, wherein the logistics evaluation data includes timeliness data, safety data, cost data, service quality data, and coverage data; S400, obtaining relevant information of the customer order, combining supplier performance data, supplier risk data and logistics performance data, to obtain the best supplier and the best logistics company matching the customer order, wherein the relevant information of the customer order includes customer order data and customer address data; S500. Based on the customer address data, an initial order delivery path is obtained through a path selection algorithm and the initial order delivery path is optimized based on a path optimization algorithm to obtain an order delivery path, and the customer order delivery is completed based on the order delivery path and the customer order data.

[0039] The present invention obtains supplier performance data, supplier risk data and logistics performance data by analyzing supplier evaluation data of suppliers and logistics evaluation data of logistics companies, thereby obtaining the best supplier and the best logistics company, and obtaining the order delivery path through the relevant information of customer orders, and completing the delivery of customer orders by combining the order delivery path and customer order data. The method of the present invention comprehensively analyzes suppliers and logistics companies, selects suppliers and logistics companies with good performance and stability for customers, realizes stable production and delivery of customer orders, and improves service quality.

[0040] In this embodiment, glass trading is taken as an example, and a glass trading platform is provided. The trading process of the glass trading platform is as follows: Figure 3 The glass trading platform obtains and analyzes the supplier evaluation data of the target supplier to obtain the supplier performance data, wherein the supplier evaluation data includes quality data, price data and delivery data. The supplier performance data is obtained through the following steps: Step 1: Obtain the target supplier's qualified rate data and customer evaluation data through customer feedback, where the customer evaluation data includes after-sales service, customer evaluation and communication efficiency. Based on the qualified rate data and customer evaluation data, obtain the quality data, which is expressed as follows:

[0041] Step 2: Obtain the supplier quotation data of the supplier, analyze the market data to obtain the market average price, and analyze the supplier quotation data and the market average price to obtain price data. The supplier with a price lower than the market average and a smaller price fluctuation obtains a higher price data. The price data is expressed as follows:

[0042] Step 3: Analyze the supplier's historical delivery data and set a score for on-time delivery rate. For example, if 100% of the delivery is on time, the on-time delivery rate is 100 points, and the on-time delivery rate is reduced based on the on-time delivery ratio to obtain the on-time delivery rate. Combined with the supplier's response to the emergency, analyze the emergency response capability data, and combine the on-time delivery rate to obtain the delivery data. The delivery data is expressed as follows:

[0043] Step 4: Establish a supplier scoring model, analyze the supplier evaluation data, obtain supplier performance data, and classify suppliers based on the supplier performance data. In this embodiment, suppliers are divided into three levels: A, B, and C. Class A suppliers (supplier performance scores above 90 points) have the best quality, price, delivery time and service performance; Class B suppliers have good performance but room for improvement; Class C suppliers need to be improved. Among them, the supplier performance data is expressed as follows:

[0044] in, Indicates quality data, Indicates the pass rate data, Represents customer evaluation data, Represents price data, Represents supplier quotation data, represents the average market price, Indicates delivery data, Indicates the on-time delivery rate, Represents emergency response capability data, represents supplier performance data, , , , Indicates the evaluation data weight, , , Indicates the weight of supplier performance data.

[0045] In order to improve the accuracy and comprehensiveness of supplier evaluation, the glass trading platform classifies or clusters supplier performance data through the supplier analysis model, and then identifies potential high-risk suppliers through classification and clustering results. At the same time, if the supplier continues to show volatility or instability in a certain data indicator, the glass trading platform sets up an early warning mechanism and adjusts the supplier risk label. In this embodiment, the supplier analysis model includes one of an SVM classification model, a K-means clustering model, and a Bayesian classification model, wherein the SVM classification pre-training model obtains an SVM classification model by training the supplier performance data and the corresponding real risk labels, and classifies and predicts the supplier performance data based on the SVM classification model to obtain supplier risk data, including the following steps: Step 1: Obtain the supplier performance data of the target supplier and conduct risk assessment to obtain the real risk label corresponding to the supplier performance data of the target supplier; Step 2: Establish an SVM classification pre-training model and a risk assessment loss function, train based on the target supplier performance data and the corresponding real risk labels, and optimize through the risk assessment loss function to obtain an SVM classification model. The risk assessment loss function is expressed as follows:

[0046] Step 3: Classify and predict the risk labels of the supplier performance data to be tested based on the SVM classification model to obtain the supplier risk data; in, represents the risk assessment loss function, Indicates the target supplier performance data, Indicates the real risk label corresponding to the supplier evaluation data, represents the regularization parameter, represents the normal vector of the SVM classification model, represents the bias of the SVM classification model, Indicates Supplier performance data, Represents the quantity of supplier performance data.

[0047] The K-means clustering model clusters the supplier performance data and analyzes the clustering results to obtain supplier risk data, including the following steps: Step 1: Preset the number of classification clusters and classification cluster centers, and perform cluster allocation based on the distance between the supplier performance data and the classification cluster center. Allocate the supplier performance data to the classification cluster closest to it, and calculate the classification cluster center after cluster allocation. The distance between the supplier performance data and the classification cluster center is calculated using the following formula:

[0048] The classification cluster center is calculated by the following formula:

[0049] Step 2: Iteratively perform cluster assignment and calculation of classification cluster centers on supplier performance data until the classification cluster centers no longer change, obtain the performance data classification results, set a distance threshold, obtain supplier performance data in the data classification results whose distance from the classification cluster centers does not meet the distance threshold, and analyze them to obtain supplier risk data; in, represents the distance between the supplier performance data and the classification cluster center, represents supplier performance data, Indicates Supplier performance data, Indicates The cluster centers, represents the quantity of supplier performance data, Indicates the number of classification clusters.

[0050] The Bayesian classification model analyzes and predicts the real risk labels and prior probabilities of supplier performance data by using the Bayesian theorem to obtain supplier risk data, including the following steps: Step 1: Obtain the prior probability of supplier risk, analyze historical supplier performance data, and obtain the prior probability of historical supplier performance data and the prior probability of supplier performance when supplier risk occurs; Step 2: Based on the prior probability of supplier risk, the prior probability of historical supplier performance data corresponding to when the supplier has risk, and the prior probability of supplier performance, analyze the supplier performance data to obtain the supplier risk category probability, which is expressed as follows:

[0051] Step 3: Select the maximum value of the supplier risk category probability as the supplier risk data, which is expressed as follows:

[0052] in, represents the supplier risk category probability, represents the prior probability of supplier risk, Represents the prior probability of historical supplier performance data corresponding to when a supplier is at risk, represents the prior probability of supplier performance, Indicates Supplier performance data, represents supplier performance data, Indicates the number of supplier performance data. Represents supplier risk data, Represents historical supplier risk data.

[0053] In this embodiment, in order to ensure the real-time nature of supplier risk data, supplier risk data is automatically updated monthly or quarterly, and time series forecasts are made based on supplier risk data, and risk forecasts are made in advance for supplier risk data for a period of time in the future, so as to help customers avoid risks in advance. In the glass trading platform of this embodiment, an evaluation report is generated for suppliers, including supplier performance data, supplier risk data, and supplier level, and a line chart is used to display the supplier's historical performance trend chart to indicate the trend of changes in each supplier's indicators over time. The supplier risk data is represented by a scatter plot, which helps customers to quickly screen each supplier.

[0054] Obtaining logistics evaluation data of logistics companies, and analyzing the logistics evaluation data based on a logistics data analysis model to obtain logistics performance data, including the following steps: Step 1: Analyze the historical order volume and delivery time of the logistics company to obtain the on-time delivery order data and total order data. The time efficiency data is obtained through analysis. A higher time efficiency data indicates that the logistics company has a strong time management ability and can meet customer needs. The time efficiency data is expressed as follows:

[0055] Step 2: Check whether the logistics company has anti-vibration and anti-collision measures during transportation, obtain the damaged order data of the logistics company, and analyze it in combination with the total order data to obtain safety data. The higher the safety data, the stronger the professionalism of the logistics company in transportation. The safety data is expressed as follows:

[0056] Step 3: Compare the quotation data of logistics companies and the average price data of logistics through the cost scoring formula to obtain cost data. The higher the cost data, the higher the cost performance of the logistics company. The cost data is expressed as follows:

[0057] Step 4: Obtain data such as customer feedback and complaint rates of logistics companies, analyze customer feedback data, service response data and tracking information data to obtain service quality data, and obtain coverage data through logistics network data and transportation coverage data; Step 5: Standardize the timeliness data, safety data, cost data, service quality data and coverage data, take the standardized timeliness data, cost data, service quality data and coverage data as the ideal solution, take the standardized safety data as the negative ideal solution, obtain the closeness of the logistics evaluation data, and obtain the logistics performance data, which is expressed as follows:

[0058] in, Indicates time-sensitive data. Indicates on-time delivery of order data, Represents the total order data, Indicates safety data. Indicates broken order data, Represents cost data, Indicates the quotation data of logistics companies. Represents the average logistics price data, Represents logistics performance data, represents a negative ideal solution, represents an ideal solution.

[0059] In this embodiment, the glass trading platform is connected to the logistics company system to track the logistics performance data in real time, including monitoring the logistics transportation process through GPS to ensure that each transportation node can arrive on time during the transportation process; if the transportation is delayed or the risk of damage is detected, the glass trading platform automatically sends an early warning; collects and analyzes real-time feedback from customers, scores the service quality of all orders, and updates the logistics performance data in real time. According to the logistics performance data, a regular assessment and elimination system is established to ensure that the logistics company maintains a high level of service. If the logistics company is below the preset threshold in the continuous scoring, the glass trading platform can suspend cooperation with the logistics company, and reward the logistics company that continues to maintain a high level of service by allocating more orders or giving priority to cooperation opportunities, so as to achieve the purpose of screening high-level logistics companies and ensure the safety, punctuality and economical transportation of products during the transportation of customer orders.

[0060] In this embodiment, the glass trading platform collects the category, brand, grade, film system and specification model information of the glass sheet supplier's products and puts them on the shelves. At the same time, the platform provides a variety of payment options, seamlessly connects with the payment gateway and third-party payment channels, and customers place orders according to their needs. The system automatically generates the deposit to be paid according to the relevant information of the customer's order. After the customer completes the deposit payment, the online submission, summary and analysis of procurement needs are realized to improve the accuracy of the procurement plan. The platform sorts the supplier performance data, supplier risk data and logistics performance data according to the relevant information of the customer's order, selects the supplier with better supplier performance data and lower supplier risk data as the best supplier, and selects the logistics company with better logistics performance data as the best logistics company, so as to achieve the purpose of reducing procurement costs and improve the customer's procurement efficiency and bargaining power. Among them, the relevant information of the customer order includes the customer order data and the customer address data.

[0061] After determining the best supplier, the glass trading platform determines the balance of the order based on the final price and triggers an automatic reminder process, such as notifying the customer of the deadline for the balance payment through email, text message or in-site message, to avoid problems with the customer's order due to forgetting the balance; the customer pays the balance of the order according to the order settlement information, and the specific process of the transaction is clearly displayed on the glass trading platform, including the customer's payment to the glass trading platform, order review, the glass trading platform's payment to the supplier, and the supplier's delivery, etc., to increase the transparency of the transaction process. After the customer pays the balance, the supplier understands the customer's specific needs and expectations based on the relevant information of the customer's order, and arranges the production sequence based on its own equipment capabilities, staffing and material supply, reducing customer waiting time and production cycle, giving priority to urgent orders and important customer orders, ensuring on-time delivery of orders, and reducing the risk of production fluctuations and inventory backlogs by analyzing customer orders and inventory levels, avoiding excess or out-of-stock situations.

[0062] After the best supplier based on customer order data completes glass production, the order is delivered based on the best logistics company. In the order delivery process, it is often the case that the vehicle returns empty after delivery, resulting in low transportation efficiency. In order to reduce the empty rate in the delivery process, the delivery route needs to be optimized. Among them, the empty rate is mainly affected by the following factors: 1. Order density and order distribution. Areas with high order density and concentrated distribution are more likely to achieve full-load delivery, while areas with dispersed orders may have a higher empty rate; 2. Seasonal and volatile demand. The demand for glass products will fluctuate with the seasonal changes in the construction and other markets, resulting in uneven transportation demand in certain time periods; 3. The degree of matching of round-trip transportation demand. If the round-trip order or other close orders can be matched on the delivery route, the return empty rate can be reduced; 4. Vehicle scheduling and resource optimization. Unreasonable scheduling of transportation vehicles will lead to an increase in the empty rate and low resource utilization; 5. Real-time road conditions and traffic conditions. In order to achieve efficient delivery of orders, it is necessary to consider the traffic conditions during the order delivery process to avoid congestion and other factors that affect transportation efficiency.

[0063] In this embodiment, the empty load rate is optimized through the following aspects: 1. Analyze customer order data and group orders with similar destinations or similar areas for delivery, that is, arrange multiple orders in the same vehicle for delivery, so as to improve the utilization rate of vehicles in a single transportation; 2. Establish a round-trip order matching mechanism, analyze the potential order demand in the order delivery path through the glass trading platform, match new orders in the return path, realize "return trip with goods", and reduce the empty load situation of the return trip; 3. Dynamically adjust the delivery plan, analyze future order demand through the demand analysis model, arrange the delivery and return paths in advance, and dynamically adjust to adapt to the fluctuating order demand, such as identifying the peak order area through the time series prediction method, and realize flexible distribution of distribution resources; 4. Layered scheduling strategy, layer customer orders according to priority, time window, etc., reasonably allocate delivery vehicles, and realize full load operation of transportation vehicles, such as arranging orders with higher priority on vehicles with abundant resources, and delivering orders with lower priority by grouping, so as to ensure efficient transportation.

[0064] Based on the relevant information of the best logistics company and the customer order, the initial order delivery path is obtained. In this embodiment, the initial order delivery path is optimized by the shortest path algorithm and the vehicle path optimization algorithm to obtain the order delivery path. The shortest path algorithm is based on the transportation heuristic function and is suitable for finding the shortest path under the condition of real-time road condition changes, and includes the following steps: Step 1: Get the transport node based on the customer order data, create a transport open list and a transport closed list based on the transport node, get the start node and the destination node through the transport node and add them to the transport open list; Step 2: Establish a transportation heuristic function, select the transportation node that minimizes the transportation heuristic function from the open list as the current node, and add the current node to the closed list. Iterate until the current node is the target node to obtain the order delivery path, where the transportation heuristic function is expressed as follows:

[0065] in, represents the transport heuristic function, Indicates the actual cost from the starting node to the current node, represents the heuristic estimate of the cost from the current node to the target node, Indicates a transport node.

[0066] The vehicle routing optimization algorithm is used to optimize the initial order delivery path. The vehicle routing optimization algorithm is used to optimize the multi-order and multi-vehicle delivery problem, which can effectively reduce the empty load rate and improve the vehicle utilization rate. According to the distance and time window of the order address data, the optimal path and load of the vehicle in a single delivery process are calculated to maximize the vehicle loading rate; the vehicle routing optimization algorithm with a time window considers the order time limit and plans the shortest time route for each transport vehicle to ensure that the delivery is completed within the specified time of the order. Among them, the vehicle routing optimization algorithm includes the following steps: Step 1: Get the transportation node based on the customer order data, obtain the order transportation cost through the transportation node, obtain the node decision to indicate whether to pass through the current transportation node, and establish the transportation objective function based on the order transportation cost and node decision, as shown below:

[0067] Step 2: By adjusting the node decision and combining the order transportation cost between transportation nodes, the transportation objective function is minimized to obtain the order delivery path; in, represents the transportation objective function, Indicates the transport node To transport node The shipping cost of orders between Indicates passing through a transport node To transport node Node decision making, Indicates the number of transport nodes.

[0068] This embodiment provides an optional path optimization algorithm, which uses a genetic algorithm to optimize the complex initial order delivery path. The genetic algorithm simulates the selection, crossover and mutation of biological evolution to search for the optimal order delivery path, which is suitable for complex delivery paths with multiple orders and multiple constraints. In the genetic algorithm, the delivery path of the customer order is encoded into a chromosome sequence, and each gene represents a delivery point; the shortest total time or the shortest total distance of the path is set as the fitness target; individuals with higher fitness are selected for crossover and mutation to obtain a new delivery path, and the order delivery path is obtained through step-by-step optimization. At the same time, the glass trading platform uses GPS and real-time traffic data to update the order delivery path in real time to reduce congestion during the order delivery process.

[0069] After completing the delivery of customer orders, this embodiment updates the logistics performance data through order delivery data, where the order delivery data includes empty rate data, delivery time efficiency data and comprehensive path optimization data. The empty rate data is obtained by calculating the empty rate of each delivery, and the logistics company with a lower empty rate has a higher score; the delivery time efficiency data is obtained by tracking the time efficiency of each delivery route; the comprehensive path optimization data is obtained by comprehensively scoring the cost, distance and empty rate of the transportation route. Through the score update, the glass trading platform can continuously optimize the selection of logistics companies, empty rate control and delivery efficiency, and achieve more efficient and economical transportation management.

[0070] Embodiment 2: An order procurement and delivery system based on data analysis, such as Figure 2 As shown, it includes a supplier evaluation module 100, a risk analysis module 200, a logistics evaluation module 300, an order confirmation module 400 and an order delivery module 500; The supplier evaluation module 100 obtains supplier evaluation data of a target supplier, builds a supplier scoring model to analyze the supplier evaluation data, and obtains supplier performance data, wherein the supplier evaluation data includes quality data, price data, and delivery data; The risk analysis module 200 establishes a supplier analysis model, performs risk assessment on supplier performance data based on the supplier analysis model, and obtains supplier risk data; The logistics evaluation module 300 obtains logistics evaluation data of the logistics company, builds a logistics data analysis model to analyze the logistics evaluation data, and obtains logistics performance data, wherein the logistics evaluation data includes timeliness data, safety data, cost data, service quality data and coverage data; The order determination module 400 obtains relevant information of the customer order, combines supplier performance data, supplier risk data and logistics performance data, and obtains the best supplier and the best logistics company that match the customer order, wherein the relevant information of the customer order includes customer order data and customer address data; The order delivery module 500 obtains the initial order delivery path based on the customer address data through a path selection algorithm and optimizes the initial order delivery path based on a path optimization algorithm to obtain the order delivery path, and completes the customer order delivery based on the order delivery path and the customer order data.

[0071] Various changes and modifications can be made without departing from the spirit and scope of the present invention, and all equivalent technical solutions also belong to the scope of the present invention.

[0072] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0073] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, devices, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0074] The present invention is described with reference to the flowcharts and / or block diagrams of the method, terminal device (system), and computer program product according to the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of the processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing terminal device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing terminal device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0075] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing terminal device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured product including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0076] These computer program instructions can also be loaded onto a computer or other programmable data processing terminal device so that a series of operating steps are executed on the computer or other programmable terminal device to produce a computer-implemented process, thereby providing instructions for executing on the computer or other programmable terminal device to implement the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0077] It should be noted that: The "one embodiment" or "embodiment" mentioned in the specification means that the specific features, structures or characteristics described in conjunction with the embodiment are included in at least one embodiment of the present invention. Therefore, the phrases "one embodiment" or "embodiment" appearing in various places throughout the specification do not necessarily refer to the same embodiment.

[0078] In addition, it should be noted that the shapes and names of the parts and components of the specific embodiments described in this specification may be different. Any equivalent or simple changes made based on the structure, features and principles described in the patent concept of the present invention are included in the protection scope of the patent of the present invention. The technicians in the technical field of the present invention can make various modifications or supplements to the specific embodiments described or replace them in a similar manner, as long as they do not deviate from the structure of the present invention or exceed the scope defined by the claims, they should all fall within the protection scope of the present invention.

Claims

1. An order purchasing and delivery method based on data analysis, characterized in that: The following steps are involved: Obtain supplier evaluation data of target suppliers, build a supplier scoring model to analyze the supplier evaluation data, and obtain supplier performance data, where the supplier evaluation data includes quality data, price data, and delivery data; Establish a supplier analysis model, conduct risk assessment on supplier performance data based on the supplier analysis model, and obtain supplier risk data; Obtain logistics evaluation data of logistics companies, build a logistics data analysis model to analyze the logistics evaluation data, and obtain logistics performance data, where the logistics evaluation data includes timeliness data, safety data, cost data, service quality data and coverage data; Obtain relevant information about customer orders, and combine supplier performance data, supplier risk data, and logistics performance data to obtain the best supplier and best logistics company that match the customer order, wherein the relevant information about customer orders includes customer order data and customer address data; Based on customer address data, the initial order delivery path is obtained through the path selection algorithm and the initial order delivery path is optimized based on the path optimization algorithm to obtain the order delivery path, and the customer order delivery is completed based on the order delivery path and customer order data.

2. The order purchasing and delivery method based on data analysis according to claim 1 is characterized in that: The step of building a supplier scoring model to analyze supplier evaluation data and obtain supplier performance data includes the following steps: Obtain and analyze the target supplier's supply qualification rate data and customer evaluation data to obtain the target supplier's quality data, where the customer evaluation data includes after-sales service, customer evaluation and communication efficiency. The quality data is expressed as follows: Obtain supplier quotation data and market average price and analyze them to obtain price data of target suppliers. The price data is expressed as follows: Obtain and analyze the on-time delivery rate and emergency response capability data to obtain the target supplier's delivery data, which is expressed as follows: A supplier scoring model is established based on quality data, price data and delivery data. The target supplier evaluation data is analyzed based on the supplier scoring model to obtain supplier performance data, which is expressed as follows: in, Indicates quality data, Indicates the pass rate data, Represents customer evaluation data, Represents price data, Represents supplier quotation data, represents the average market price, Indicates delivery data, Indicates the on-time delivery rate, Represents emergency response capability data, represents supplier performance data, , , , Indicates the evaluation data weight, , , Indicates the weight of supplier performance data.

3. The order purchasing and delivery method based on data analysis according to claim 1 is characterized in that: The supplier analysis model includes one of an SVM classification model, a K-means clustering model and a Bayesian classification model; The SVM classification model is trained by fitting the supplier performance data and the corresponding real risk labels, and performs classification prediction based on the training results and the supplier performance data to obtain the supplier risk data; The K-means clustering model clusters the supplier performance data and analyzes the clustering results to obtain supplier risk data; The Bayesian classification model analyzes and predicts the real risk labels and prior probabilities of supplier performance data through Bayesian theorem to obtain supplier risk data.

4. The order purchasing and delivery method based on data analysis according to claim 3 is characterized in that: The SVM classification model is trained by fitting the supplier performance data and the corresponding real risk labels, and performs classification prediction based on the training results and the supplier performance data to obtain the supplier risk data, including the following steps: Conduct risk assessment on supplier performance data and obtain the real risk label corresponding to the supplier performance data; Establish an SVM classification pre-training model and a risk assessment loss function, train the SVM classification pre-training model based on supplier performance data and the corresponding real risk labels, and optimize it based on the risk assessment loss function to obtain the SVM classification model; Based on the SVM classification model, the performance data of the supplier to be tested is risk assessed to obtain the supplier risk data; The risk assessment loss function is expressed as follows: in, represents the risk assessment loss function, represents supplier performance data, Indicates the actual risk label corresponding to the supplier performance data, represents the regularization parameter, represents the normal vector of the SVM classification model, represents the bias of the SVM classification model, Indicates Supplier performance data, Represents the quantity of supplier performance data.

5. The order purchasing and delivery method based on data analysis according to claim 3 is characterized in that: The K-means clustering model clusters the supplier performance data, analyzes the clustering results, obtains outlier data and analyzes them to obtain supplier risk data, including the following steps: The number of classification clusters and the classification cluster centers are preset, cluster allocation is performed based on the distance between the supplier performance data and the classification cluster centers, and the classification cluster centers after cluster allocation are calculated, wherein the distance between the supplier performance data and the classification cluster centers is calculated by the following formula: The classification cluster center is calculated by the following formula: Iterate the cluster allocation and the calculation of the classification cluster center until the classification cluster center no longer changes, obtain the performance data classification result, set the distance threshold, obtain the supplier performance data of the performance data classification result whose distance from the classification cluster center does not meet the distance threshold and analyze it to obtain the supplier risk data; in, represents the distance between the supplier performance data and the classification cluster center, represents supplier performance data, Indicates Supplier performance data, Indicates The cluster centers, Indicates the quantity of supplier performance data, Indicates the number of classification clusters.

6. The order purchasing and delivery method based on data analysis according to claim 3 is characterized in that: The Bayesian classification model analyzes and predicts the real risk labels and prior probabilities of supplier performance data through the Bayesian theorem to obtain supplier risk data, including the following steps: Obtain the prior probability of supplier risk, analyze historical supplier performance data, and obtain the prior probability of historical supplier performance data and the prior probability of supplier performance when supplier risks occur; Based on the prior probability of supplier risk, the prior probability of historical supplier performance data corresponding to when the supplier has risk, and the prior probability of supplier performance data, the supplier performance data is analyzed to obtain the supplier risk category probability, which is expressed as follows: The maximum value of the supplier risk category probability is selected as the supplier risk data, which is expressed as follows: in, represents the supplier risk category probability, represents the prior probability of supplier risk, Represents the prior probability of historical supplier performance data corresponding to when a supplier is at risk, represents the prior probability of supplier performance data, Indicates Supplier performance data, represents supplier performance data, Indicates the number of supplier performance data. Represents supplier risk data, Represents historical supplier risk data.

7. The order purchasing and delivery method based on data analysis according to claim 1 is characterized in that: The construction of the logistics data analysis model to analyze the logistics evaluation data to obtain logistics performance data includes the following steps: Obtain the on-time delivery order data and total order data of the logistics company, and analyze them to obtain the timeliness data, which is shown as follows: Get the damaged order data of the logistics company, analyze it in combination with the total order data, and get the safety data, which is shown as follows: Obtain and analyze the quotation data of logistics companies and the average price data of logistics to obtain the cost data, which is shown as follows: Analyze customer feedback data, service response data and tracking information data to obtain service quality data, and obtain coverage data through logistics network data and transportation coverage data; The timeliness data, safety data, cost data, service quality data and coverage data are standardized, and the standardized timeliness data, cost data, service quality data and coverage data are taken as the ideal solution, and the standardized safety data is taken as the negative ideal solution. The closeness of the logistics evaluation data is obtained, and the logistics performance data is obtained, which is expressed as follows: in, Represents time-sensitive data, Indicates on-time delivery of order data, Represents the total order data, Indicates safety data. Indicates broken order data, Represents cost data, Indicates the quotation data of logistics companies. Represents the average logistics price data, Represents logistics performance data, represents a negative ideal solution, represents an ideal solution.

8. The order purchasing and delivery method based on data analysis according to claim 1 is characterized in that: The optimization of the initial order delivery path includes the following steps: Analyze customer order data and customer address data, group customer orders with similar destinations for delivery, and perform return delivery based on customer order data; Establish a demand forecasting model, combine customer order data to predict customer order demand, obtain the priority ranking of customer order data, and dynamically adjust the initial order delivery path based on customer order demand and priority ranking; The initial order delivery path is optimized through the shortest path algorithm and vehicle path optimization algorithm to obtain the order delivery path.

9. The order purchasing and delivery method based on data analysis according to claim 8 is characterized in that: The optimization of the initial order delivery path by the shortest path algorithm includes the following steps: Get the transport node based on the customer address data, create a transport open list and a transport closed list based on the transport node, get the start node and the destination node through the transport node and add them to the transport open list; Establish a transportation heuristic function, select the transportation node that minimizes the transportation heuristic function from the open list as the current node, and add the current node to the closed list. Iterate until the current node is the target node to obtain the order delivery path, where the transportation heuristic function is expressed as follows: in, represents the transport heuristic function, Indicates the actual cost from the starting node to the current node, represents the heuristic estimate of the cost from the current node to the target node, Indicates a transport node.

10. The order purchasing and delivery method based on data analysis according to claim 8, characterized in that: The optimization of the initial order delivery path by the vehicle path optimization algorithm includes the following steps: Based on the customer order data, the transportation node is obtained, the order transportation cost is obtained through the transportation node, and the node decision is obtained to analyze whether to pass through the current transportation node. The transportation objective function is established based on the order transportation cost and node decision, which is expressed as follows: By adjusting the node decision and combining the order transportation cost between transportation nodes, the transportation objective function is minimized and the order delivery path is obtained; in, represents the transportation objective function, Indicates the transport node To transport node The shipping cost of orders between Indicates passing through a transport node To transport node Node decision making, Indicates the number of transport nodes.

11. The order purchasing and delivery method based on data analysis according to claim 1, characterized in that: It also includes updating the logistics performance data through order delivery data after completing the delivery of customer orders, where the order delivery data includes empty load rate data, delivery time data and comprehensive route optimization data; The empty load rate data is obtained by obtaining the empty load rate of the best logistics companies in the process of customer order delivery and analyzing it; The delivery time efficiency data is obtained by real-time tracking of the time efficiency of the best logistics company during the delivery process and analyzing the time efficiency of reaching the transportation node during the transportation process; The route optimization data is obtained through a comprehensive analysis of the cost, distance and vehicle empty rate of each transportation route of the best logistics company.

12. An order purchasing and delivery system based on data analysis, characterized in that: It includes supplier evaluation module, risk analysis module, logistics evaluation module, order confirmation module and order delivery module; The supplier evaluation module obtains supplier evaluation data of the target supplier, builds a supplier scoring model to analyze the supplier evaluation data, and obtains supplier performance data, wherein the supplier evaluation data includes quality data, price data, and delivery data; The risk analysis module establishes a supplier analysis model, performs risk assessment on supplier performance data based on the supplier analysis model, and obtains supplier risk data; The logistics evaluation module obtains logistics evaluation data of the logistics company, builds a logistics data analysis model to analyze the logistics evaluation data, and obtains logistics performance data, wherein the logistics evaluation data includes timeliness data, safety data, cost data, service quality data and coverage data; The order determination module obtains relevant information of the customer order, combines supplier performance data, supplier risk data and logistics performance data, and obtains the best supplier and the best logistics company that match the customer order, wherein the relevant information of the customer order includes customer order data and customer address data; The order delivery module obtains the initial order delivery path based on the customer address data through a path selection algorithm and optimizes the initial order delivery path based on a path optimization algorithm to obtain the order delivery path, and completes the customer order delivery based on the order delivery path and the customer order data.

13. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 11 is implemented.

14. An order purchasing and distribution device based on data analysis, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the method according to any one of claims 1 to 11 is implemented.

Images